International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

Assessment of Transformer Oil Quality Using Fuzzy Logic Technique

INTRODUCTION The life of a transformer is dependant upon life of its insulation. The insulation of a transformer deteriorates over a span of time with temperature, moisture and oxygen. Oil quality assessment is the technique to decide the quality of transformer oil after a specific period of use. The quality of transformer oil decides the efficiency of the transformer while in service. This chapter presents an innovative methodology for assessment of transformer oil quality by using fuzzy logic based decision support system. The quality of transformer oil can be termed as good if it fulfills the following functional requirements for at least 30 years of service. It acts as an electrical insulator. It acts as a coolant for transformer winding and core. Transformer oil free from water particles and sludge acts as a good insulator. Gases are produced in a transformer when it is subjected to electrical, thermal and environmental stresses. These gases get dissolved in oil and provide information about the incipient faults developing in transformer. The oil quality assessment is necessary for deciding preventive maintenance schedules of transformers. So far during the course of study of literature, no work has been found on the assessment of oil quality for deciding future course of action, after a specific period of use. In this dissertation work, an attempt is being made to assess oil quality by using fuzzy logic to arrive at a decision regarding the action to be taken.

IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

RECONDITIONING OF TRANSFORMER OIL There is a tendency of transformer oil to absorb water vapour from atmosphere due to breathing during service and even during transportation and storage. Water is injurious to transformer insulation system due to the following reasons [1]. It reduces the electric strength of oil. It reduces the resistivity of oil. It accelerates deterioration of solid insulation. Air dissolved in oil induces risk of bubble formation and accelerated oxidation process, leading to chemical deterioration of oil. The biggest contaminator, moisture enters transformer oil from following sources. by leakage past gasket by absorption within the transformer as a product of degradation of insulation at high temperatures Transformer oil also gets impregnated with various types of impurities. These may be solid ones like hygroscopic fibers, suspended particles or liquid ones like dissolved water organic and inorganic gases. They bring about a considerable reduction in the dielectric strength of the oil. As an example, oil at 20oC saturated with water (44 PPM) attains only about 25% of the original electric strength with water content of 10 PPM. Thus, transformer oil is reconditioned to eliminate the following elements. free and dissolved water solid impurities dissolved gases The possible reconditioning processes are: Single Filtering and Degassing Double Filtering and Degassing Reclamation

IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

The oil quality assessment method developed in this dissertation is based on finding the Total Hydrocarbon Gas (TCG) content in the oil. Depending upon TCG and severity of different gases present, the Oil Degradation Index (ODI) is evaluated and used to assign one of the above possible courses of action, if it is suitable for reconditioning. The single stage and double stage filtering and degassing are done by vacuum plants. This procedure involves course and fine filtration followed by degassing and dehumidification. The transformer oil is spread over a large surface area under vacuum, after which oil is delivered for use. The single stage and double stage filtering decision depends upon degradation level of used oil as manifested by the concentration of dissolved gases. In addition there are two other classifications. No Filtering Do Not Use No Filtering condition means oil is not degraded and can be reused without any requirement of reconditioning. Do Not Use condition means whole transformer oil is to be replaced with new one, as the oil is degraded to an extent that its reconditioning has become impossible. Oil reclamation is a process to remove heterogeneous atoms and oxidized products from oil. This is a chemical treatment of oil followed by filtering and degassing. The aim is to remove acid and other harmful compounds by precipitation, sludge formation or by use of chain reaction of organic compounds.

FUZZY MEASURE FOR ASSESSMENT OF TRANSFORMER OIL QUALITY It is easy to visualize that considerable uncertainty is involved in the process of defining transformer oil quality. Hence, it is imperative to use fuzzy logic for this purpose. The fuzzy measure for the assessment of oil quality is
IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

based on a combination of knowledge about TCG along with particular key gases evolved due to fault condition. The decision making process is based on the degree of match between current and permissible limits of dissolved gases and the fuzzy rule based system. The method is designed to compute the fuzzy measure named as transformer Oil Degradation Index (ODI), by integration of the information of different hydrocarbon gases formed in transformer oil. This approach uses a knowledge base to derive Oil Degradation Index (ODI). The transformer oil quality data is transformed into a normalized fuzzy number with membership grade function adjusted for characterizing transformer oil quality based on the data available from both normal and faulty transformers.

Fuzzy Membership Functions The input variables to the present classification have different ranges in various hydrocarbon gases viz. CH4, H2, C2H6, C2H4, and C2H2 developed during incipient fault detection on the basis of DGA and TCG (Total Gas Content). Trapezoidal membership functions (trapmf) are used for all input variables. The term set of membership functions for all input variables except TCG are: Low Moderate High Severe The input variable TCG uses another term set V. High in addition to above term sets. The input membership functions along with different term sets are shown in Figure 4.1. This figure is generalized for different inputs. The ranges of membership functions for input variables are shown in Table 4.1. There is only one output variable to assign transformer oil quality. This is called Oil Degradation Index (ODI). ODI value is computed for each input
IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

variable defined above. This ODI value and severity of fault as defined by rule base, decide the process to be adopted for reconditioning. The input membership functions are designed to give a constant output over substantial range. The membership of each input variable varies over a definite range, uniformly on both sides (trapezoidal). The full membership (value 1) covers 80% area of total range. This is the reason why output gives a constant ODI in maximum cases thus, making this measure insensitive to small changes in gas concentrations. This is an extremely useful property to take care of normal changes due to operations and ageing. Another significant feature is that this ODI is further correlated to the specific reconditioning procedure of transformer oil. The output membership functions are shown in Figure 4.2 and corresponding range and type of membership functions are shown in Table 4.2.

input i Low 1 Moderate High V. High Severe

input i

Fig. : Input membership function used in oil reconditioning

IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

ODI MF 1 1 MF 2 MF 3 MF 4 MF 5

1 ODI

Fig. : Output membership function used in oil reconditioning Note: MF 1 = MF 2 = MF 3 = MF 4 = MF 5 = No Filtering Single Filtering & Degassing Double Filtering & Degassing Reclamation Do Not Use

IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

Fuzzy Rules The fuzzy rules are developed to match a specific output condition of oil depending upon various input values. 38 rules are developed in all. Table 4.3 shows the input data for oil quality analysis. The corresponding outputs are shown in Table 4.4. The output oil quality or Oil Degradation Index (ODI) is shown for individual group of inputs. Table 4.3: Input Data for Oil Quality Assessment
INPUT DATA C2H6 4 1 0 3 4 4 3 5 3 8 4 2 13 3 0 16 54 17 16 67 4 5 1 27

S. No. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

CH4 4 2 2 3 2 4 3 20 4 10 15 10 7 9 1 125 212 107 31 144 95 20 32 117

H2 3 1 2 2 2 2 2 9 7 14 13 13 17 12 0 95 68 32 42 60 1076 240 338 531

C2H4 3 0 0 1 3 3 0 10 1 5 3 1 3 7 0 281 470 265 115 449 71 28 32 132

C2H2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 231 96 50 848

TCG 14 4 4 9 11 13 8 44 15 37 35 26 40 31 1 517 804 421 204 729 1477 389 453 1655

IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.

71 38 104 248 262 91 43 1107 47 1417 4 99 122 339 4 34 21 5365 61 87 186 38 16615 1393 770 8784 13 584 10 619 3997 24 4066 1053 695 207 38 1393 770 8784 1742 95

79 41 110 125 222 98 48 201 442 114 2 747 41 59 7 21 199 2973 65 16 813 212 2754 800 199 4906 24 266 160 80 231 127 9474 507 416 441 212 800 199 4906 425 1076

3 3 30 10 27 18 3 110 117 296 3 13 31 42 3 5 0 427 16 75 15 15 3657 304 217 1404 5 328 3 326 1726 0 353 297 74 43 47 304 217 1404 7299 4

72 45 86 147 168 72 75 2016 67 2096 4 97 143 392 2 47 40 5532 143 395 249 47 31476 2817 1508 9924 43 862 1 2480 5584 32 6552 1440 867 224 15 2817 1508 9924 37043 71

115 58 131 179 410 154 81 6350 69 0 0 1036 188 1 0 62 144 2124 3 30 1001 78 613 3000 72 9671 319 1 1 0 0 81 12997 17 0 261 78 3000 72 9671 158 231

340 185 461 709 1089 433 250 9784 742 3923 13 1992 525 833 16 169 404 16421 288 603 2264 390 55155 8314 2766 34689 404 2041 175 3505 11538 264 33442 3314 2050 1176 390 7314 2766 34689 62349 1477

IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78.

754 167 1324 369 27 370 20 79 22 144 9739 107

244 117 858 137 274 1249 240 33 307 60 2004 127

172 48 208 144 5 56 5 30 2 67 2750 11

1281 481 2793 1242 33 606 28 215 33 449 5113 157

27 7 7672 16 97 1371 96 5 109 9 0 224

2478 820 12855 1908 436 3652 389 362 473 729 19606 623

Table 4.4: OUTPUT: OIL Degradation Index and Suggested Reconditioning Method
S. No. Oil Degradation Index (ODI) 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.0784 0.2500 0.2500 0.2500 0.2500 0.2500 0.4500 0.2500 0.2500 0.4500 Reconditioning Method No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering No Filtering Single Filtering and Degassing Single Filtering and Degassing Single Filtering and Degassing Single Filtering and Degassing Single Filtering and Degassing Double Filtering and Degassing Single Filtering and Degassing Single Filtering and Degassing Double Filtering and Degassing
IJSER 2012 http://www.ijser.org

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

10

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72.

0.2500 0.2500 0.3500 0.2500 0.3500 0.2500 0.2500 0.8610 0.4500 0.6500 0.0784 0.4500 0.4500 0.2500 0.0784 0.2500 0.2500 0.7770 0.2500 0.4500 0.6500 0.2500 0.7770 0.7770 0.6500 0.7770 0.4500 0.4500 0.2500 0.6500 0.7770 0.2500 0.7770 0.6500 0.4500 0.3500 0.2500 0.7770 0.4500 0.7770 0.8710 0.4500 0.4500 0.2500 0.7770 0.4500 0.2500 0.6500

Single Filtering and Degassing Single Filtering and Degassing Double Filtering and Degassing Single Filtering and Degassing Double Filtering and Degassing Single Filtering and Degassing Single Filtering and Degassing Do Not Use Double Filtering and Degassing Reclamation No Filtering Double Filtering and Degassing Double Filtering and Degassing Single Filtering and Degassing No Filtering Single Filtering and Degassing Single Filtering and Degassing Do Not Use Single Filtering and Degassing Double Filtering and Degassing Reclamation Single Filtering and Degassing Do Not Use Do Not Use Reclamation Do Not Use Double Filtering and Degassing Double Filtering and Degassing Single Filtering and Degassing Reclamation Do Not Use Single Filtering and Degassing Do Not Use Reclamation Double Filtering and Degassing Double Filtering and Degassing Single Filtering and Degassing Do Not Use Double Filtering and Degassing Do Not Use Do Not Use Double Filtering and Degassing Double Filtering and Degassing Single Filtering and Degassing Do Not Use Double Filtering and Degassing Single Filtering and Degassing Reclamation
IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

11

73. 74. 75. 76. 77. 78.

0.2500 0.2500 0.2500 0.2500 0.7770 0.2500

Single Filtering and Degassing Single Filtering and Degassing Single Filtering and Degassing Single Filtering and Degassing Do Not Use Single Filtering and Degassing

4.2

CONCLUSION The method applied in this work is original and innovative in nature.

During periodical incipient fault diagnosis of transformers, Oil Degradation Index (ODI) gives additional information for representing quality of oil. This helps in better monitoring of transformer condition. The output results of oil quality assessment can be used for making appropriate decisions regarding the procedure to be adopted for the reconditioning of transformer oil. This will improve life of transformer oil and life of transformer as a whole. Hence, preventive maintenance schedules can be programmed optimally with minimum breakdown.
REFERENCES

1.

Bharat Heavy Electricals Limited, Bhopal, Transformers, Tata McGrawHill Publishing Company Limited, New Delhi, 1997. 2. A.C. Franklin and D.P. Franklin, The J&P Transformer Book, 11th Edition, Aditya Book Pvt. Ltd, New Delhi 1995. 3. Joseph B. Digiorgio, Dissolved Gas Analysis of Miniral Oil Insulating Fluids, http.//www.nttworldwide.com/tech2102.htm. 4. J.A. Momoh, X.W. Ma, K. Tomsovic, Overview and Literature Survey of Fuzzy Set Theory in Power System. IEEE Transactions on Power Systems, Vol. 10, N0-3, August 1995. 5. K.Tomsovic, A.Amar, On Refining Equipment Condition Monitoring Using Fuzzy Sets and Artificial Neural Nets, International Journal of Engineering Intelligent Systems, Vol.5, No.1, March 1997, 43-50.

IJSER 2012 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 ISSN 2229-5518

12

6. G.N.S. Kalyani D.V.S.S. Sivasarma, AI Techniques for Condition Monitoring of Power Trasformers Using DGA. National Power Systems Conference, NPSC 2004, 1110 1115. 7. A.G. Levit, O.N. Grechko, N.P. Shchipunova, Allowance for Insulation Ageing in the New Concept of Accelerated Life Tests of High Voltage Power Transformers, IEEE Transactions on Power Delivery, Vol.7, No.3, July 1992. 8. R.M. Radwan, R.M. El-Dewieny, I.A. Metwally, Investigation of Static Electrification Phenomenon Due to Transformer Oil Flow in Electric Power Apparatus,IEEE Transaction on Electrical Insulation, Vol. 27, No. 2, April 1992. 9. IEEE Substations Committee Working Group K4, Partial Discharge Testing of Gas Insulated Substations, IEEE Transactions on Power Delivery, Vol.7, No.2, April 1992. 10.J Aubin, Y. Langhame, Effect of Oil Viscosity on Transformer Loading Capability at Low Temperatures, IEEE Transactions on Power Delivery, Vol.7, No.2, April 1992. 11.W.J. Mcnutt, T.O. Rouse, G.H. Kaufmann, Mathematical Modelling of Bubble Evolution in Transformers, IEEE Transaction on Power Apparatus and Systems, vol. 104, No. 2, February 1985.

IJSER 2012 http://www.ijser.org